Enhanced endogenous gene tagging in human induced pluripotent stem cells via AAV6-mediated donor delivery

Abstract

Systematically tagging endogenous proteins with fluorescent markers in human induced pluripotent stem cells (hiPSCs) allows observation of live cell dynamics in different cell states. However, the precise insertion of fluorescent proteins into live cells via CRISPR/Cas9-induced editing relies on homology-directed repair (HDR). The nonhomologous end-joining (NHEJ) DNA repair pathway often outcompetes HDR, resulting in irreversible insertions and deletions (INDELs) and low knock-in efficiency. Recognizing successful HDR-mediated tagging events is an additional challenge when the target gene is not expressed in stem cells and successful tagging cannot be immediately observed. To address these challenges, we used: 1) adeno-associated virus serotype 6 (AAV6) mediated DNA donors at optimized multiplicity of infection (MOI) to deliver tag payloads at maximal efficiency; 2) titrated, multiplexed Cas9:gRNA ribonucleo-protein (RNP) amounts to assure balanced HDR/INDEL frequency among conditions; 3) long-amplicon droplet digital PCR (ddPCR) to measure the frequency of HDR-generated alleles in edited pools; and 4) simultaneous Inference of CRISPR Edits (ICE) to detect and thereby avoid conditions significantly saturated (>50%) with INDELs. These approaches enabled us to identify efficient and accurate editing conditions and recover tagged cells, including cells tagged at loci not expressed in stem cells. Together these steps allowed us to develop an efficient methodology and workflow to clonally isolate directly from an ideal cell pool with optimal HDR and minimized INDEL frequencies. Using this approach, we achieved both monoallelic and biallelic insertion of fluorescent markers into four genes that are turned on during differentiation but not initially expressed in hiPSCs, where direct selection of tagged cells based on fluorescence was impossible:TBR2, TBXT, CDH2(pro-differentiation and pro-migratory genes), andCDH5(endothelial specific gene). Through a systematic evaluation of various gRNA sequences and RNP concentrations, we identified conditions for each gene that achieved high HDR frequencies, peaking at 38.6%, while also avoiding conditions saturated with INDELs, where isolation of clones with a tagged allele in trans with an unedited allele is difficult. Over-all, this methodology enhances the efficiency of fluorescent tag knock-in at genes not expressed in hiPSCs, facilitating reliable image-based observation of cellular processes, and enables recovery of accurately edited mono- and biallelically tagged clones. We standardized these approaches to yield an efficient and general workflow for introducing large HDR mediated knock-ins into hiPSCs.